The Best (Expected) Secondary Pitches of 2019

Yesterday, I put every fastball thrown in baseball last year into a giant spreadsheet to come up with expected pitch values. Well fine, it was a small snippet of code, not a giant spreadsheet. But the output came in a giant spreadsheet! In any case, the idea is pretty straightforward: look at a player’s pitches, substitute in xwOBA-based contact numbers instead of actual results, and call it a metric.

Today, I’m completing the set. Well, I’m kind of completing the set; I ignored knuckleballs because there aren’t enough of them, and secondary offerings are more complex. Due to differing classification systems, I scraped breaking balls (sliders, curveballs, knuckle curves, and even cutters) and offspeed pitches as a single pitch type. Otherwise, we might end up with something like Nick Anderson — classification systems can’t decide if he throws a curve or a slider.

One more thing: the system is heartless. No human could argue that this wasn’t the best curveball of the year:

Or if not that one, then this one, with bonus Eric Lauer bewilderment and Greinke sprinting:

Slow curves are undoubtedly the best curves, results be damned. But the soulless calculation robot doesn’t agree with me on that, caring about “whether the opposition hit it” and “whether it gets strikes” instead of “whether Ben audibly giggles when the pitch is thrown.” To each their own, I suppose.

Ignoring aesthetic value, here are the top 10 breaking balls by expected value:

What do you know? Patrick Corbin is on there after all. Sure, it’s mainly for his slider, but let’s ignore that for the moment. The top of this list makes sense, and it’s heartening for Nats fans to see that Scherzer deserved better while Corbin and Strasburg were as good as advertised.

Aside from that, there are some wiggly parts to breaking balls that come from pitch classification. For an example, look no further than our list of underperformers:

Leclerc, per Pitch Info, barely throws any sliders. But if you use other databases, a lot of his changeups are classified as sliders. Some of his placement at the top of this list comes down to the fact that he has 490 tracked breaking balls in this accounting, while Pitch Info thinks he only threw 40.

That aside, he also got really unlucky! When opponents put these pitches into play, they produced an xwOBA of .282. That’s quite good for production on contact (eighth percentile); he was jamming people left and right. They produced an actual wOBA of .459, which is the 85th percentile. Our metric is capturing that unluckiness, in addition to some pitch classification weirdness.

The other side of the coin shows off something else about expected values; namely, that stadium matters:

The Giants dominate this list, and that’s partially because xwOBA overestimates production in pitcher’s stadiums. Bumgarner and Samardzija also throw cutters, and cutter/four-seamer/slider differentiation is always a bear to unravel.

Lastly, we’ve got the worst expected breaking balls. These pitchers shouldn’t be throwing these pitches as much as they do:

Michael Wacha’s curveball and cutter will be particularly frustrating to Cardinals fans (and likely to Mets fans in 2020), but there are a lot of pitchers on here who would benefit from de-emphasizing their bendy pitches.

We’ll quickly run through the same exercise for offspeed pitches before I cover a few takeaways from the whole exercise. The best ones are mostly names you’d expect:

Luis Castillo really does destroy people with his change piece. And Kirby Yates is on there as a reliever, which is wild. The pitchers these stats like more than traditional pitch values are mostly guys with good changeups who had down years, and I’ll admit to being happy that the results are intuitive:

The luckiest offspeed pitchers run the gamut from “are we sure this metric works” to “yeah I could see it”:

Zach Davies was tremendously lucky on changeups that opponents put in play last year. Greinke and Ryu are sad, but not unbelievable, names to see here — their changeups still grade out well, just not as well as before. And the rest are mostly guys who I didn’t think of as having particularly good offspeed pitches but who put up good numbers in 2019.

The worst expected offspeed pitches bring us full circle back to Trevor Williams, my example from yesterday’s article:

But really, there aren’t any completely grim offspeed pitches in the same way that there are awful breaking balls and fastballs. That’s because if you have a bad changeup, you mostly just don’t throw it. You can’t get away with not throwing a fastball, and few pitchers can completely ignore breaking balls, but offspeed? That’s more of a luxury.

Finally, let’s talk caveats. The first one is this: don’t use this data as some proof of pitch greatness. There are all kinds, and I do mean all kinds, of issues with this approach. Pitch classification is tricky. Separating out good xwOBA from bad xwOBA conclusions is tricky. Understanding the role of pitches working together? You guessed it — it’s tricky.

Additionally, one of the downsides of this project from my perspective is the link with our extant pitch values. As I’m pulling from a different pitch database than Pitch Info, plenty of the discrepancies come from different pitches being included in each sample rather than the naive assumption about a difference between “actual” and “expected” value would imply.

So don’t take these ratings as gospel. Don’t compare them to the data we already provide and think you’ve cracked the code. But as a novelty, they’re great. Want to know if Jack Flaherty’s breaking balls outstripped his fastball when it comes to this value metric? Well, they did! Want to know if Yu Darvish’s array of ridiculous breaking balls is as fun as you think? It is! Want to know if Glenn Sparkman should stop throwing curves? He should!

As long as you’re using these as a fun tool, you can’t go too wrong. And there’s some chance they’re more useful from a predictive lens than our old pitch values, something I plan on exploring in the coming weeks. Or you could explore it — the data is here. I hope you enjoyed this look into how well various pitches should have done in 2019. And if you didn’t — well, I hope the slow curve GIFs make up for it.